On Tue, 16 Dec 2008, John Salter wrote (on eprints-tech):

      Hi,
      We're looking at some issues with ePrints, one of which
      is:
      'We have a metadata-only record in our repository, for
      which the
      full-text is freely available online (e.g. in arXiv). How
      should this be
      logged in the 'full_text_status' field in ePrints?'

      The full_text_status currently allows values of:
      None
      Public
      Restricted

      If we added an extra option of 'external_archive' (or
      similar), would it
      cause any problems?
      Does ROAR/OpenDOAR use this field at all (e.g. for ROAR's
      estimate of
      full text coverage)?

      How does any one else deal with this scenario?

      Cheers,
      John


Dear John, if I may make a policy-related rather than a technical
reply:

The best thing to do is to use SWORD to import the institution's
externally deposited documents and metadata, so as to make the IR
complete. Arxiv is the easiest external repository to do this from.

The only three options for a document are absent, OA, or Restricted
(Closed).

During the time you are setting up the SWORD protocol to import, if
you
already have the metadata in the IR, but not the document, then
insert
the link to the external URL, but still set the access as OA (or
public,
or what have you). This will help the growth of universal OA far more
than tagging an item non-OA just because it is located
institution-externally rather than institution-internally.

ROAR so far only tracks presence/absence of full-text approximately.
I
hope that by the time the capability of tracking full-text-status
exactly is available, this issue will have been resolved, with all
available institutional full-texts automatically imported. But if
not,
then the presence of the link to the OA full-text should be counted
as
the existence of the OA full-text document, so that the global OA
tally
(for which the locus of the OA fulltext in a network of
OAI-interoperable
IRs is less important than its existence) can be correctly estimated.

ROARMAP is a registry of mandates rather than repositories or
repository
contents, so the question does not apply to ROARMAP.

Best wishes,

Stevan Harnad
American Scientist Open Access Forum
http://amsci-forum.amsci.org/archives/American-Scientist-Open-Access-Forum.h
tml

Chaire de recherche du Canada           Professor of Cognitive
Science Institut des sciences cognitives        Electronics &
Computer Science
Universite du Quebec a Montreal         University of Southampton
Montreal, Quebec                                Highfield,
Southampton
Canada  H3C 3P8                                 SO17 1BJ United
Kingdom
http://www.crsc.uqam.ca/              
 http://users.ecs.soton.ac.uk/harnad/




Reply via email to